Abstract Silymarin production by hairy root culture of Milk this-tle (Silybum marianum L. Gaertn) was investigatedusing Agrobacterium rhizogenes AR15834. Hairy rootswere induced by injection or inoculation of explantswith A. rhizogenes. One month old hairy roots weredissected from the explants and grown in Murashingand Skoog (MS) liquid medium. Polymerase ChainReaction (PCR) using the Β gene and the Β-glucoro-nidase (GUS) assays were used for identification ofthe transformed hairy roots. Flavonolignan levels in thehairy roots were determined by high-performance liq-uid chromatography (HPLC). Five different compo-nents were isolated; taxifolin, silychristin, silydianin,silybin and isosilybin, with the following quantities,0.009, 0.041, 0.042, 0.007 and 0.011 mg g-1 dry weih-ht, respectively. Silybin was the major flavonolignan.Produced by hairy roots culture may serve as a usefulsystem for producing silymarin or studying its biosyn-thetic pathways. Keywords: Silybum marianum; Agrobacterium rhizo-genes; Hairy root; Flavonolignans.

Silymarin is an antihepatotoxic polyphenolic substance

isolated from the milk thistle plant, Silybum marianum,

(Kvasnicka et al., 2003). Silymarin consists of a large

number of flavonolignans, including silybin (SBN),

isosilybin (ISBN), silydianin (SDN), silychristin (SCN)

and taxifolin (TXF) as precursor of silymarin

(Sanchez-sampedro et al., 2005a and Hasanloo et al.,

2005b). Silymarin and silybin have been so far mostly

used as hepatoprotectants (Sonnenbichler et al., 1999).

These metabolites have also been shown to have other

interesting functions, such as anticancer and cancero-

protective properties, neurodegenerative and neurotox-

ic repressing activities. They are also associated with

the treatment and prevention of gastrointestinal prob-

lems, nephropathy, cardio-pulmonary problems and

skin protection (Van Erp et al., 2005; Veladimir et al.,2005; Katiar, 2002). Silymarin compounds are usually

extracted from dried fruits of field grown plants that

often require months to years to obtain.

The in vitro production of plant secondary metabo-

lites can be possible through plant cell culture under

controlled conditions and free from environmental

fluctuations. However, the major limitations of cell

cultures are their instability during long-term culture

and low product yields (Bonhomme et al., 2000).

Therefore great efforts have been focused on trans-

formed hairy roots (Kim et al., 2002). Hairy roots, the

results of genetic transformation by Agrobacteriumrhizogenes (Hu and Du, 2006) have attractive proper-

ties for secondary metabolite production, as compared

to differentiated cell cultures (Dhakulkar et al., 2005;

Kim et al., 2002). Root induction is due to the integra-

tion and subsequent expression of a portion of bacter-

ial DNA (T- DNA) from the bacterial Ri (Root induc-

ing) plasmid in plant genom. Four loci involved in root

formation have been identified in the T- DNA of the Ri

plasmids and designated root loci (rol) A, B, C and D.

(Ayala-Silva et al., 2007). Hairy roots are genetically

stable and not repressed during the growth phase of the

IRANIAN JOURNAL of BIOTECHNOLOGY, Vol. 6, No. 2, April 2008

113

Short Communication

Silymarin production by hairy root culture of Silybummarianum (L.) Gaertn

Hassan Rahnama1, Tahereh Hasanloo*2, Mohammad Reza Shams1, Roshanak

Sepehrifar2

1Department of Tissue Culture and Gene Transformation, Agricultural Biotechnology Research Institute of Iran,P.O. Box 31535-1897, Karaj, I.R. Iran 2Department of Physiology and Proteomics, Agricultural BiotechnologyResearch Institute of Iran, P.O. Box 31535-1897, Karaj, I.R. Iran

*Correspondence to: Tahereh Hasanloo, Ph.D.

Tel.: +98 261 2709652; Fax: +98 261 2704539E-mail: thasanloo@abrii.ac.ir

it’s culture (Bourgaud et al., 1999). They often grow as

fast as or faster than plant cell cultures (Srivastava and

Srivastava, 2007). The greatest advantage of hairy

roots is that their cultures often exhibit approximately

the same or greater biosynthetic capacity for secondary

metabolite production compared to their mother plants

(Kim et al., 2002).

Hairy root culture of S. marianum could therefore be

an alternative method for the production of flavonolig-

nanas, however, untill now, few studies have been con-

ducted in this field, with very low production observed

in the fruits of this plant (Alikaridis et al., 2000). In this

study, we describe an efficient protocol for development

of the hairy root culture of S. marianum and production

of silymarin, by using A. rhizogenes.

Dried fruits (Hungarian seeds) derived from milk

thistle were supplied by the Institute of Medical Plants

(Karaj, Iran) and the Iranian Academic Center for

Education, Culture and Research (ACECR). The seeds

were cultured in hormone-free MS (Murashige and

Skoog, 1962) medium and incubated in the dark at 26

± 1ºC for a photoperiod of 16 h light (Hasanloo et al.,2008). The bacterial strain A. rhizogenes AR15834 (a

gift from Dr P. Noroozi, the Sugar Beet Seed Institute,

Karaj, Iran) harbouring (UK) or devoid of the binary

vector pBI121 (Clontech, Canada) were used in hairy

root induction. The binary vector pBI121 carries genes

such as nptll coding for neomycin

phosphotransferaseII driven by the nopaline synthase

promoter, and an uidA gene coding for β-glucoro-

nidase (Gus) driven by a cauli flower mosaic virus

(CaMV) 35S prompter. Various excised explants such

as the hypocotyl, leaf and cotyledons were isolated

from in vitro grown seedlings. All explants were

precultured for 3 days on hormone-free medium con-

taining MS salts, vitamins and 3% (w/v) sucrose. The

medium was adjusted to pH 5.8 before adding agar (7

g/l). The precultured explants were immersed in

overnight cultures of bacterial suspension (optical den-

sity at 600 nm, (OD600 = 0.7) for 10 min and then blot-

ted dry on sterile filter paper and incubated under light

in the same medium. After three days, they were trans-

ferred to a medium containing MS salts and vitamins,

3% (w/v) sucrose, 250 mg/l cefotaxime and 7 g/l agar.

Within 4- 5 weeks, roots emerged from the wounded

sites. In all cases where A. rhizogenes harboring

pBI121was used, the medium was supplemented with

50 mg/l kanamycin. In the second approach, the hairy

roots were induced by injection of the overnight cul-

ture of A. rhizogenes suspension culture (OD600 = 0.7)

into the hypocotyls of the whole seedling using an

insulin syringe. Hairy roots which arose mainly from

the cut surface of the explants, after reaching a length

of 4-5 cm, were separated and subcultured in the dark

on the same media explained above. Wild type root

cultures were established in hormone-free MS liquid

medium as control. Hairy root lines were maintained

by transfer of 3-4 long root pieces to hormone- free

liquid medium containing MS salts, vitamins and 3%

(w/v) sucrose at 25ºC on a rotary shaker (130 rpm) in

complete darkness and subcultured every 2 weeks.

Root genomic DNA was extracted as described by

Khan et al. (2007). Polymerase chain reaction (PCR)

was performed in 35 thermal cycles (denaturation at

94ºC for 1 min, primer annealing at 53ºC for 1 min,

and primer extension at 72ºC for 1 min) for rolB(Forward primer 5´-ATGGATCCCAAATTGC-

TATTCCCCACGA-3´ and Reverse primer 5´- TTAG-

GCTTCTTTCATTCGGTTTACTGCAGC-3´) and in

35 thermal cycles (denaturition at 94ºC for 1 min,

primer annealing at 61ºC for 1 min, and extension at

72ºC for 2 min) for GUS (Forward primer 5´-

GGTGGGAAAGCGCGTTACAAG-3´ and Reverse

primer 5´-TGGATTCCGGCATAGTTAAA-3´) gene

specific primers. The β-glucoronidase histochemical

assay was performed according to the method by

Jefferson et al. (1987) using root segments. Roots were

immersed in sodium phosphate buffer (pH 7.0) con-

taining 2 mM 5- bromo-4-chloro-3- indolyl-β-glucoro-

nic acid. The reaction was allowed to proceed for 20 h

in the dark at 37ºC (Jefferson et al., 1987). The one

month old hairy roots were harvested from the liquid

medium and placed between the folds of a blotting

paper to remove excess water and were then freeze-

dried. Root growth was measured in terms of dry

weight (DW). The samples were treated with ethyl

acetate to remove fats. The flavonolignans were

extracted from the dried residue with 10 ml of

methanol at 40ºC for 8 h. The methanolic solution was

concentrated under vacuum to a dry residue and re-dis-

solved in 2 ml of methanol and kept at 4ºC in the dark

(Cacho et al., 1999). Flavonolignans levels were deter-

mined by high-performance liquid chromatography

(HPLC) according to the method by Hasanloo et al.(2005a). This procedure involved the use of a Knauer,

Germany liquid chromatography system equipped

with a Knauer injector consisting of a 20 µl loop, a

Eurosphere (Knauer, Germany) C185 µ (250 × 4.6

Rahnama et al.

114

mm) column, a Knauer K2600A UV detector with

Chromgate software for peak integration. The mobile

phase consisted of the solvents; acetonitrile: water

(40:60) with 10% (v/v) H3PO4, (pH 2.6). All solvents

and chemicals were of HPLC grade (Sigma, Aldrich,

Germany). The elution time and flow rate were 30 min

and 1 ml/min respectively, and the resulting peaks

were detected at 288 nm. Identification was achieved

by comparison of the sample retention times (Rt) with

those of the standards SCN, SDN, SBN, TXF and a

standard mixture of SLM.

The flavonolignan content of the root was expressed

as mg/g (roots DW) and derived using a known con-

centration of standard and sample peak areas. The data

obtained from the analysis of each sample allowed the

plotting of a calibration curve showing good linearity

(a correlation coefficient of 0.999). The data were dis-

played as the mean of at least three replicates.

Statistical significance was calculated using the

Duncan test for unpaired data (α′≥0.05) and the

Analysis of variance (ANOVA) method was used for

comparisons of the means. Statistical analysis was

made by Statistical Analysis Software (SAS) software

(Version 6.2). Standards of SLM, SBN and TXF were

purchased from Sigma (Germany); SCN and SDN

were obtained from Phytolab (Germany).

Hairy root cultures are usually able to produce the

same compounds that can be found in vivo, without the

loss of concentration frequently observed with callus

or cell suspension cultures. A few studies addressing

the possibility of flavonolignan production in “invitro” cultures have been carried out and in all cases

the production has been found to be very low and even

disappearing during prolonged cultures (Namdeo,

2007; Tumova et al., 2004).

In the present study, we observed that hairy root

induction in S. marianum can be made by inoculation

of hypocotyl, cotyledon and leaf explants with A. rhi-zogenes AR15834. In all the explants (cases), hairy

roots were induced in 7-10 days after inoculation,

emerging on the wounded side of the explants (Fig. 1).

In the first experiment for optimizing hairy root trans-

formation, the efficiency of transgenic root selection

based on screening of hairy-roots for GUS activity was

compared in explants of S. marianum. Of 150 cotyle-

don explants inoculated with A. rhizogenese contain-

ing the pBI121 vector, 48 roots were produced after 4

weeks. Subsequent histochemical GUS staining of root

tissues confirmed GUS activity in 45 (30%) of the

hairy root clones. Transformation efficiencies were

7.9% for hypocotyls, 21.6% for cotyledons and 20%

for whole plants by using the injection method. All of

the Gus positive hairy roots as tested by histochemical

analyses were confirmed by PCR analyses of the rolBand gus transgenes (Fig. 2).

Because of simple handling, possibility of less con-

IRANIAN JOURNAL of BIOTECHNOLOGY, Vol. 6, No. 2, April 2008

115

Figure 1. Hairy root induction of S. marianum. leaf (1), hypocotyl

(2), whole plant via the injection method (3), and cotyledon (4), Hairy

roots in liquid culture (5). GUS staining in hairy roots indicates suc-

cessful integration of β-glucoronidase transgene (6).

Figure 2. PCR analysis of hairy roots for gus (left) and rolb (right) transgenes. 1: Molecular weight mark-

er, 2: Non transgenic root, 3: Positive control, 4- 6: Transgenic hairy roots.

tamination and higher transformation efficiency, we

therefore, propose inoculation of cotyledon and leaf

explants for induction of hairy roots in S. marianum. In

the second experiment cotyledon explants were trans-

formed using A. rhizogenes without the reporter gene,

for induction of hairy roots. Transformed hairy roots

were selected via PCR analysis of the rolB gene.

Eight different hairy root lines were established on

liquid MS medium and compared analytically with

non-transformed roots. Significant statistical differ-

ences were observed among the lines. The highest bio-

mass production was found in line 7 (about 1 g). The

Rahnama et al.

116

Figure 3. Chromatogram (HPLC) of methanolic extract of hairy roots of S. marianum.

Three flasks were harvested for each root line; 1-8: Hairy root lines; 9: untransformed roots; ND: non-detectable;. S.D: standard deviation. The superscript

letters following the calculated means and standard deviations are an indication of similarity in the data. Values sharing a letter within a column are not sig-

nificantly different at p< 0.05.

Table 1. Dry weight (DW) and flavonolignans contents (mg/g DW) (± SD) produced by S. marianum hairy root lines after 4 weeks in

culture. The flavonolignans were determined by HPLC method.

Isosilybin Silybin Silychristin Silydianin Taxifolin D W (g)

0.001 ± 0.000 c 0.002 ± 0.001 de 0.012 ± 0.001 g 0.019 ± 0.001 d 0.015 ± 0.001 b 0.953 ± 0.008b

0.005 ± 0.003 bc 0.004 ± 0.002 bcd 0.030 ± 0.001 d 0.091 ± 0.001 a 0.019 ± 0.001 a 0.367 ± 0.010f

0.008 ± 0.001 b 0.006 ± 0.001 ab 0.072 ± 0.004 b 0.067 ± 0.002 b 0.004 ± 0.000 ef 0.178 ± 0.025g

ND ND 0.141 ± 0.005 a 0.086 ± 0.003 a 0.002 ± 0.000 e 0.057 ± 0.006h

0.011 ± 0.001 a 0.007 ± 0.001 a 0.041 ± 0.002 c 0.042 ± 0.004 c 0.009 ± 0.000 c 0.352 ± 0.014f

0.006 ± 0.001 bc 0.005 ± 0.000 abc 0.029 ± 0.002 d 0.026 ± 0.002 d 0.012 ± 0.07 b 0.638 ± 0.006d

0.003 ± 0.001 bc 0.004 ± 0.001 bcd 0.022 ± 0.003ef 0.022 ± 0.002 d 0.015 ± 0.005 b 1.000 ± 0.011a

0.002 ± 0.000 bc 0.003 ± 0.002 dc 0.024 ± 0.004 g 0.022 ± 0.002 d 0.013 ± 0.001 b 0.853 ± 0.011c

0.003 ± 0.001 bc 0.002 ± 0.001 de 0.018 ± 0.002 f 0.019 ± 0.003 d 0.006 ± 0.000 cd 0.554 ± 0.056e

lines were also assessed for the production of flavono-

lignans. The flavonolignan content varied greatly from

one line to another. The presence of silybin and isosily-

bin were detected by HPLC analysis of the methanolic

extract of the hairy root culture sample (Fig. 3).

Further experiments will be conducted on the single

hairy root line showing the highest biomass and

flavonolignans production.

We have shown that such roots developed on S.

marianum produced SBN (0.007 mg/g DW), ISBN

(0.011), SCN (0.041), SDN (0.042) and TXF (0.009)

similar to those produced by dried fruits of this plant

(Table 1, line 5). In this case, while accumulating in

the seeds, silymarin has also been shown to accumu-

late in the hairy root cultures. This is in agreement

with some previously reported results in other plants

(Kim et al., 2002), although the yield percentage is

lower than what is generally found in dried fruits of

S. marianum. The major flavonolignans in the

untransformed root culture were SBN (0.002 mg/g

DW), ISBN (0.003), SCN (0.018), SDN (0.019) and

TXF (0.006).

The hairy root cultures of S. marianum can be a

promising source for continuous and standardized

production of silymarin under controlled conditions.

Only one report has been published regarding hairy

root induction and flavonolignan production in S.

marianum cultures. In that report silymarin produc-

tion was demonstrated to be very low and the accu-

mulation of silybin was not even detected (Alikaridis

et al., 2000). Transformed root cultures of several

other species have been evaluated for their content of

secondary metabolites relative to other wild type

plants. The secondary product profiles have often

been observed to be conserved. For example hairy

roots of ginseng (Panax ginseng Meyer) produced

the same saponins and ginsenoides as the wild type

roots. (Yoshikawa et al., 1987). Hairy root culture has

been used as a model system to study various aspects

of the metabolic and molecular regulation of several

secondary metabolites. For example, Chen et al.(2000) have shown that methyl viologen, a generator

of the superoxide anion triggers the formation of

cryptotanshione in hairy root cultures of Salvia milti-orrhiza. Savita et al. (2006) have studied the effects

of several biotic and abiotic elicitors in hairy root cul-

tures of Beta vulgaris in the shake-flask and bioreac-

tor. They were able to show a good level of betalanin

production (1.2% or 88.4 mg/l) using the hairy root

culture system.

The availability of this protocol for the production

of silymarin similar to those described above provides

a powerful system to study various aspects of the

metabolic and molecular regulation of silymarin

biosynthesis. It is also necessary to evaluate and screen

the effects of various elicitors with different mecha-

nisms on the production and accumulation of sily-

marin for pharmaceutical industries.

Acknowledgments

We thank Dr. P. Noroozi for providing A. rhizogenes strain

used in this work. This research was funded by the

Agricultural Biotechnology Research Institute of Iran (ABRII).

References

Alikaridis F, Papadakis D, Pantelia K, Kephalas T (2000).

Flavonolignan production from Silybum marianum transfor-

med and untransformed root culture. Fitotrapia. 71: 379-384.

Ayala-Silva T, Bey CA, Dortch G (2007). Agrobacterium rhizoge-nez mediated transformation of Asimina triloba L. Cuttings.

Pakistan J Bio Sci. 10: 132-136.

Bonhomme V, Laurain-Matter D, Lacoux J, Fliniaux MA and

Jacquin-Dubreuil A (2000). Tropan alkaloid production by

hairy roots of Atropa belladona obtained after transformation

with Agrobacterium rhizogenes 15834 and Agrobacteriumtumefaciens containing rol A, B, C genes only. J Biotechnol.81: 151-158.

Bourgaud F, Bouque V and Guckert A (1999). Production of

flavonoids by Psoralea hairy root cultures. Plant Cell TissOrg. 56: 97-104.

Cacho M, Moran M, Corchete P, Fernandez-Tarrago J (1999).

Influence of medium composition on the accumulation of

flavonolignans in cultured cells of Silybum marianum (L.)

Gaertn. Plant Sci. 144: 63-68.

Chen H, Chen F (2000). Induction of phytoalexin formation in

crown gall and hairy root cultyre of Salvia miltiorrhiza by

methyl viologen. Biotechnol Lett. 22: 715- 720.

Dhakulkar S, Bhargava S, Ganapathi TR, Bapat VA (2005).

Induction of hairy roots in Gmelina arborea Roxb. Using

Agrobacterium rhizogenes. Founder’s Day Special Issue:

100-106.

Hasanloo T, Khavari-Nejad RA, Majidi E, Shams-Ardakani MR

(2008). Flavonolignan production in cell suspension culture

of Silybum marianum (L.) Gaertn. Pharm Biol. (In press).

Hasanloo T, Khavari-Nejad RA, Majidi E, Shams-Ardakani MR

(2005a). Analysis of flavonolignans in dried fruits of Silybummarianum (L.) Gaertn from Iran. Pakistan J Bio Sci. 8: 1778-

1782.

Hasanloo T, Khavari-Nejad RA, Majidi E, Ziai SA, Shams-

Ardakani MR (2005b). Determination of flavonolignan of

dried fruits of Silybum marianum L. Gaertn collected from

different areas of Iran by spectrophotometer, TLC and HPLC.

IRANIAN JOURNAL of BIOTECHNOLOGY, Vol. 6, No. 2, April 2008

117

J Med Plant. 4: 25-32.

Hu ZB, Du M (2006). Hairy root and its application in plant genet-

ic engineering. J Int Plant Biol. 48: 121-127.

Jefferson RA, Kavanagh TA, Bevan MW (1987). GUS fusions: β -

Glucuronidase as a Sensitive and Versatile Gene Fusion

Marker in Higher Plants. EMBO. 6: 3901-3907.

Katiyar SK (2002). Treatment of silymarin, a plant flavonoid, pre-

vents ultraviolet light-induced immune suppression and

oxidative stress in mouse skin. Int J Oncl. 21: 1213-1222.

Khan IUH, Gannon V, Kent R, Koning WR, Lapen D, Miller J,

Neumann N, Phillips R, Robertson W, Topp E, Bochove EV,

Edge TA (2007). Development of a rapid quantitative PCR

assay for direct detection and quantification of culturable and

non-culturable Escherichia coli from agriculture watersheds.

J Microbiol Methods. 69: 480-488.

Kim Y, Wyslouzil B, Weathers PJ (2002). Secondary methabolism

of hairy root cultures in bioreactors. In Vitro Dev Biol Plant.38: 1-10.

Kvasnicka F, Biba B, Sevcik R, Voldrich M, Kratka J (2003).

Analysis of the active components of silymarin. JChromatogr A. 990: 239-245.

Murashige T, Skoog F (1962). A revised medium for rapid growth

and bioassays with tobacco tissue cultures. Physiol Plant.15:

473-497.

Namdeo AG (2007). Plant cell elicitation for production of second-

ary metabolites: A review. Phcog Rev. 1: 69-79.

Sanchez-Sampedro MA, Fernandez-Tarago J, Corchete P (2005a).

Yeast extract and methyl jasmonate induced silymarin pro-

duction in cell culture of Silybum marianum L. Gaerth. JBiotechnol. 119: 60-69.

Savita BC, Thimmaraju R, Bhagyalakashmi N, Ravishankar GA

(2006). Different biotic and abiotic elicitors influence beta-

lanin production in hairy root cultures of Beta vulgaris in

shake- flask and bioreactor. Process Biochem. 41: 50- 60.

Sonnenbichler J, Scaleram F, Sonnenbichler I, Weyhenmeyer R

(1999). Stimulatory effects of silibinin and silichristin form

the Milk thistle Silybum marianuna on kidney cells. JPharmacol Exp Ther. 29: 1375-1383.

Srivastava S, Sivastava AK (2007). Hairy root culture for mass-

production of high-value secondary metabolites. Crit RevBiotechnol. 27: 29-43.

Tumova L, Gallova K and Rimakova J (2004). Silbum marianum4in vitro. Ceska Slov Farm. 53: 135-40.

Van Erp NPH, Baker S, Zhao M, Rudek MA, Guchelaar HJ,

Nortier JWR, Sparreboom A and Gelderblom H (2005). Effect

of milke thistle (Silybum marianum) on the pharmacokinetics

of irino tecan. CLIN Cancer Res. 11: 7800-7806.

Veladimir K, Walterova D (2005). Silybin and silymarin new

effects and applications. Biomed. Papers. 149: 29-41.

Yoshikawa T, Furuya T (1987). Saponin production by cultures of

Panax ginseng transformed with Agrobacterium rhizogenes.

Plant Cell Rep. 6: 449-453.

Rahnama et al.

118